Method for updating cross section shape

There is a choice between raising or lowering the entire section (method 1), updating the regions of the section that are below water (method 2) or using a function related to local depths of flow (method 3). The choice of method should be based on observed changes to cross section shape if such observations are available.  All cross sections are treated alike, so it may be necessary to select the method that represents most of the channel being modelled. 

Method 1 results in the quickest model run times and can be used for initial model runs.  It is appropriate for channels where the whole section is under water during the entire period of simulation, except for steep banks.  Method 2 is a special case of method 3 (the exponent set to 0.0).  An indication of the effect of varying the exponent is seen in Figure 19, a case with deposition in a narrow trapezoidal channel. In the absence of any data on cross section variations method 3 with an exponent of 0.5 can be used.

The value of bed level change at which updating is undertaken should initially be set at 0.001m. Ideally, tests should be made to check that there is no reduction in accuracy due to the value chosen, and to assess the impact on run times.  Comparisons should be made against a fully accurate run with the value set to 0.0.  If method 1 is used for updating the bed, then the value 0.0 can be used as there will be little penalty on run times.

Figure 19. Comparison of bed updating methods